Association of ZIP Code Sociodemographic Characteristics with Radiation Oncology Services, Payments, and Technique Utilization

Patrick Travis Courtney^*Myung Shin SimMichelle E EalaEulanca Y LiuYa-Chen Tina ShihMichael Lee SteinbergLuca Faustino VallePuja S VenkatMichael XiangAnn Raldow

• University of California, Los Angeles, Los Angeles, United States

Introduction: Variation in radiation oncology care by provider and geography have been reported though predominantly analyzed radiotherapy delivery. We evaluated variations in radiation oncologist practices and payments by ZIP code-level sociodemographic data to broaden our understanding of these variations. Methods: We linked the CMS Medicare Physician and Other Practitioner and National Neighbor Data Archive databases to measure the association between radiation oncologist services, payments, and radiotherapy technique utilization in 2023 and the sociodemographic characteristics of their practice location ZIP code. We included all non-facility radiation oncologists for services and payments, and non-facility radiation oncologists who submitted radiotherapy treatment delivery Healthcare Common Procedure Coding System codes for technique utilization. We calculated the percentages of 3D conformal radiotherapy (3DCRT), intensity modulated radiotherapy (IMRT), stereotactic radiosurgery/stereotactic body radiotherapy (SRS/SBRT), proton radiotherapy (PRT), and brachytherapy (BT) use. We used univariable linear regression to measure the association between services, payments, and technique utilization and ZIP code-level sociodemographic variables, and multivariable linear regression to control for ZIP code-level cancer type proportions. Results: For services and payments, we identified 2,431 radiation oncologists from 1,126 ZIP codes. In univariable analyses, radiation oncologists in ZIP codes with higher education and income submitted significantly fewer services per beneficiary, as well as fewer unique services per beneficiary for the latter. Increasing ZIP code socioeconomically disadvantaged and Hispanic/Immigrant population were associated with significantly greater unique services per beneficiary in both univariable and multivariable analyses. Few associations between payments and sociodemographic factors were identified. For technique utilization, we identified 1,400 radiation oncologists from 773 ZIP codes. Radiation oncologists in ZIP codes with a predominant Non-Hispanic White population, higher education, or higher income utilized more SRS/SBRT, PRT, and BT, whereas those in socioeconomically disadvantaged ZIP codes utilized more 3DCRT. Discussion: This study provides insight into existing inequities in radiation oncology care and suggests that disparities are not limited to care delivery, but exist in radiotherapy technology access, service intensity, and to a lesser degree, payments, and may be shaped by the sociodemographic context of practice location. These data may serve as a benchmark for measuring the impact of anticipated policy changes on radiotherapy utilization.